Introduction
In the fast-evolving and complex landscape of modern healthcare, nurses are often regarded as the “eyes and ears” of care organizations [1]. They play a critical role in providing daily care to patients, especially in long-term care environments [2, 3]. Positioned on the frontlines, nurses possess a unique perspective that allows them to detect inefficiencies, identify (potential) issues or risks, and recognize opportunities for improvement in both patient care and the broader operational effectiveness of their teams or organizations [4‐6]. To transform these observations into actionable steps and meaningful organizational benefits, it is crucial for nurses to exhibit voice behavior—that is, to actively express their ideas and concerns about work-related matters to individuals capable of implementing the necessary changes [7].
Voice behavior, as an intentional, “planned behavior” occurring in an interpersonal context, not only fosters subordinates’ sense of agency but also maximizes their proactive contributions [8]. Research has highlighted voice behavior as a critical factor influencing patient safety outcomes as well as driving team and organizational improvements within care settings [5, 6, 9]. From any perspective, voice behavior has gained increasing recognition as a cornerstone of safe healthcare practices, as the failure to speak up can compromise the timely implementation of patient safety measures in hospitals [10].
Anzeige
Recent studies have demonstrated that supervisors greatly influence subordinates’ voice behavior. For example, researches have explored the effect of authentic leadership, inclusive leadership, trust in leader, abusive supervision and other factors on subordinates’ voice behavior [11‐20]. These studies have proved the mediating role of conscientiousness, psychological capital, career resilience and impression management motivation [11‐20]. Despite this progress, we believe that current approaches to examining voice can be improved in several ways.
First, previous research has overlooked the role that emotions play in supervisor-subordinate interactions. Especially in the nursing environment, nurses are subjected to a greater workload on a daily basis and may accumulate more negative emotions, which need to be attended to by their superiors. Affective leadership, as a type of leadership proposed based on the Chinese context, this type of supervisor will not only give emotional care to subordinates, but also pay attention to the emotional changes of subordinates and regulate the subordinates’ emotions, which has gradually attracted the attention of researchers. Second, previous studies have mostly explored the mechanism of supervisors’ influence on subordinates’ voice behaviors based on a cognitive perspective, with less attention paid to interpersonal interactions between supervisors and subordinates. As an important interaction object of subordinates in the workplace, supervisors not only affect the cognitive state of subordinates, but also have an impact on the interpersonal relationship between supervisors and subordinates, therefore, it is necessary to explore the influence mechanism of supervisors on subordinates’ voice behavior from the perspective of interpersonal interaction. Third, the interaction between individuals does not occur in a vacuum, the interaction between subordinates and supervisors occurs within the team, the group climate will have an impact on the subordinate behavior. In current study, we will also explore the effect of group climate.
This paper explores the factors that influence the voice behavior of nurses to provide a theoretical grounding for the effective management of nurses. We are interested in how leadership style, in particular, affective leadership, can influence the willingness of nurses to speak out and voice their ideas, suggestions and opinions to their supervisors in a bid to improve current conditions or practices in their workplace [21]. Affective leadership has been described by Weng et al. (2016) as involving “benevolence, righteousness and wisdom” qualities that pertain to the traditional Chinese Confucian cultural system [22]. We contend that it is a style of leadership that is more akin to Chinese traditions and culture and, as such, affective leadership should have a higher adaptability to Chinese organizational contexts. Because Chinese society is relationship-oriented, the supervisor-subordinate Guanxi is likely to have a profound impact on subordinates’ behavior [23]. Given this, we draw on social exchange theory and investigate how affective leadership influences the willingness of nurses to voice, taking into account the mediating role of superior-subordinate Guanxi. However, the exchange between individuals does not occur in a vacuum, the exchange between subordinates and supervisors occurs within the team, so the group climate will have an impact on the behavior of subordinate. Thus, we will also explore the effect of group voice climate.
Literature review and hypotheses
Social exchange theory (SET) is one the most influential theories in explaining workplace behaviors. It essentially involves interaction among actors with social exchanges involving resources [24]. These exchanges involve at least two actors that are dependent on each other. In workplace, such exchanges may occur between the organization, supervisors and subordinates, where any goodwill gesture by any actor may be reciprocated by other. Such reciprocity sometimes goes beyond the initial action and may show behaviors that are voluntary [25]. The caring and benevolent behavior by the supervisor in general creates an atmosphere of goodwill and subordinate may in turn form good relationships with supervisor. Based on the reciprocity principle of social exchange, subordinates incur an obligation to reciprocate their superiors. When a subordinate identifies problems or areas that can be improved in the organization, the subordinate will actively offer his or her suggestions and countermeasures in return for the supervisor. Besides, the exchange between subordinates and supervisors occurs within the team, the group climate will have an impact on the behavior of subordinate. In current study, we would like to explore the effects of affective leadership on subordinates’ voice behaviors and further explore the mediating role of superior-subordinate Guanxi and the moderating role of group voice climate.
Anzeige
Leadership behavior is one of the important factors affecting the willingness of employees to voice [7]. Positive styles of leadership, such as participative leadership [26], moral leadership [27, 28], and servant leadership [19, 29], have been shown to have a positive influence on employees’ voice behavior. Affective leadership falls into this category, prioritizing the interpersonal relationship between superiors and subordinates over the authority of leaders [22]. Affective leaders seek to show care and support to their subordinates when communicating and interacting with them. They are also inclined to actively lend a helping hand when subordinates encounter difficulties in their work or in their personal lives.
The reciprocity principle of social exchange theory suggests that, when subordinates perceive the support and care of their leaders, they will seek to return the favor to the leader or the organization. This may take the form of subordinates actively offering suggestions when they identify aspects that can improve the status quo of the organization [30] or voice when they perceive potential risks. Affective leadership empowers subordinates to voice because subordinates feel cared for, well supported, and sufficiently safe to speak their minds. Affective leaders typically have high emotional intelligence, and can not only manage their own emotions, but also those of their subordinates [22], providing the foundations for a positive culture and climate. It is not surprising, then, that studies on affective leadership and voice behavior have found a significant positive relationship between affective leadership and voice behavior [31, 32]. Based on this line of reasoning, we hypothesize that:
Hypothesis 1
Affective leadership is positively related to subordinates’ voice behavior.
The supervisor-subordinate Guanxi (SSG) in Chinese culture can be understood as a special informal social relationship established between the supervisor and the subordinate through social interactions outside of the organization [33]. It is also one of the labels of employees in organizations [34], SSG differs to the concept of leader member exchange (LMX) in western literature [35‐38]. Leader-member exchange involves only work-related forms of communication within the vertical structure of the organization [39], whereas supervisor-subordinate Guanxi extends to social interactions and communication between the supervisor and subordinate beyond the workplace [34, 35, 40]. Our contention is that affective leaders will have developed close relationships with their subordinates and, based on the theory of social exchange, this will compel subordinates to reciprocate the support they have received and to voice if it presents an opportunity to benefit the organization. Certainly, subordinates who have enjoyed a good supervisor-subordinate Guanxi will also get ample resources and support from their leaders, which will enhance their motivation to “pay back” their leader and organization [41]. Based on this logic, the findings of a meta-analysis of supervisor-subordinate Guanxi confirm a significant positive correlation between supervisor-subordinate Guanxi and employees’ voice behavior [42]. Based on this line of reasoning, we hypothesize that:
Hypothesis 2
Supervisor-subordinate Guanxi mediates the relationship between affective leadership and subordinates’ voice behavior.
Group voice climate is the collective cognition of team members on the success and safety of putting forward suggestions in the team [43], including the belief in the safety of group voice behavior and the effectiveness of group voice behavior. Morrison et al. (2011) contend that the climate within a team will influence the behaviors of individual members and their willingness to voice, and they coin the term “group voice climate” to refer to the team’s shared beliefs about voice [43]. Frazier (2009) and Frazier and Bowler (2015) found that team performance is affected by the degree to which individual members perceive the team as a safe space for actively contributing their ideas and perspectives [43‐45]. When employees believe they are in a supportive space, they believe there is a possibility that their suggestions may be given due consideration and implemented within the organization, they will be more inclined to contribute suggestions to advance the interests of the organization [43, 46, 47]. Under this condition, when subordinates and supervisors form a good supervisor-subordinate Guanxi, subordinates have reciprocal motivation, at this time, if the subordinate believes that his or her own suggestions are likely to be adopted by the supervisor, the subordinate will be more likely to make the voice behavior in return for the supervisor. Conversely, subordinates who lack this belief, or who perceive that their workplace is not a safe space to voice, the act of which may even bring negative repercussions to themselves, will be inclined to remain reticent. Although the subordinate may have a good relationship with their supervisor, this may not be sufficient to lead the subordinate to risk voicing. Based on this line of reasoning, we hypothesize that:
Hypothesis 3
Group voice climate moderates the relationship between supervisor-subordinate Guanxi and voice behavior, such that the relationship is more positive when group voice climate is higher (vs. lower).
We also contend that group voice climate can affect the indirect effect of supervisor-subordinate Guanxi on the relationship between affective leadership and voice behavior, and form a moderated mediation model. Thus, to complete our theoretical model, we predict a pattern of moderated mediation among our study variables as shown in Fig. 1.
Hypothesis 4
Group voice climate moderates the indirect effect of affective leadership on subordinates’ voice behavior through supervisor-subordinate Guanxi such that the indirect effect is stronger when group voice climate is higher.
Fig. 1
Research Model
×
![]()
Method
Sample and procedures
A questionnaire was issued to the nursing departments of three hospitals in a central province in China. Prior to issuing the questionnaire, we identified the directors of the respective hospital nursing departments to brief them on our study and aims. Following their agreement for their organization to be involved in our study, the respective directors selected a number of head nurses and their subordinates across various departments to participate in the research and to provide them with the relevant information. After consulting with them, they agreed to participate in the study. The questionnaires were distributed through online forms and printed forms. Participants were informed of the research objectives, data collection methods, and the confidentiality with which their responses would be treated. In particular, we emphasized to participants that the person in charge of their department had no ability to obtain the research data, allowing them to be candid with their responses.
This study adopted a multi-point-in-time and multi-source research paradigm. Paper questionnaires and response links were sent to participants at two time points to collect data, with a one-month interval. At Time 1, a questionnaire was issued to nurses in the department. Nurses were required to provide their demographic information and evaluate the affective leadership of their head nurses, the supervisor-subordinate Guanxi between them, and climate of the team. At Time 2, a questionnaire was issued to the head nurses who were asked to provide their demographic information and to evaluate the voice behaviors of their subordinate nurses.
Anzeige
In the first round of surveys at Time 1, 400 paper questionnaires were distributed to nurses, and 309 were effectively recovered, with an effective rate of 77.25%. At Time 2, 309 questionnaires were distributed and 236 questionnaires were recovered from the head nurses. Following the process of matching questionnaire responses between the head nurses and their subordinates, 167 matched responses were left, an effective recovery rate of 70.76%. Of the final matched responses, the mean age of the nurses in our study was 31.66(SD = 4.81); the mean age of the head nurses in our study was 41.22(SD = 5.42). With respect to education levels, the vast majority of nurses, 91.62% had a bachelor’s degree while the remaining 8.38% had a junior college education; the vast majority of the head nurses, 96.30% had a bachelor’s degree while the remaining 3.70% had master’s degree or higher. Tenure-wise, 26.95% of nurses had worked for less than 5 years, 31.74% had worked for between 6 and 10 years, 29.94% had worked for between 11 and 15 years, and 11.38% had worked for more than 15 years; 3.70% of the head nurses had worked for less than 5 years, 22.22% had worked for between 11 and 15 years, and 74.07% had worked for more than 15 years. Slightly more than half the nurses (51.50%) had worked with their head nurse for less than 5 years, 31.14% for between 6 and 10 years, 13.77% for between 11 and 15 years, and 3.59% for more than 15 years.
Measures
Each variable was measured using a 5-point Likert scale ranging from 1 “strongly disagree” to 5 “strongly agree”. Part of the measurement tools used in this study were in English, so a double-blind “translation-back” procedure was used to translate the English version of the measurement tools. The wording of some questionnaire items was adjusted to reflect the specific context of a healthcare delivery setting. Additionally, we included three attention check items (e.g., “Please answer strongly agree”) at Time 1 and Time 2. Respondents who missed such an item were immediately removed from the study.
Affective leadership
The affective leadership scale was completed by nurses and we adopted the scale developed by Weng et al. (2016) which comprised 12 items divided into two dimensions of emotional care and emotional management [22]. A sample item is “My nurse manager will take the initiative to understand the difficulties of his subordinates’ work.” Higher ratings indicate higher levels of affective leadership among head nurses. Cronbach’s α for emotional care was 0.924, 0.960 for emotional management, and 0.959 for affective leadership overall.
Voice behavior
Completed by head nurses, we used the scale developed and revised by Liang et al. (2012) which contained ten items, and which included two dimensions: promotive voice behavior and prohibitive voice behavior [21]. A sample item is “The subordinate will initiate ideas and suggestions for improving the workflow of the unit.” Higher ratings indicate more voice behavior by the subordinate. Cronbach’s α for promotive voice behavior was 0.969, 0.941 for prohibitive voice behavior, and 0.962 for the voice behavior scale overall.
Anzeige
Supervisor-subordinate Guanxi
Developed by Law et al. (2000), the supervisor-subordinate Guanxi scale was completed by nurses and included seven questions [34]. A sample item is “During holidays or after office hours, I would call my supervisor or visit him/he.” Higher ratings indicate better Guanxi between supervisors and subordinates. Cronbach’s α was 0.767.
Group voice climate
The 12-item scale by Morrison et al. (2011) was completed by nurses and included two dimensions: group voice behavior safety belief and group voice behavior effectiveness [43]. A sample item is “We can effectively advise on procedures for new projects or reforms.” Higher ratings indicate a more favorable climate for voicing. Cronbach’s α for the group voice behavior safety belief dimension was 0.930, 0.928 for the group voice behavior effectiveness dimension, and 0.958 for the group voice climate scale overall.
Control variables
Research has shown that a statistics certain correlation exists between leader member exchange and supervisor-subordinate Guanxi. Therefore, leader member exchange was taken as control variable in this study. Leader member exchange was measured through a single dimension scale comprising seven items [48]. A sample item is “Regardless of how much formal authority the position of nurse manager has, the head nurse will (probably) use his/her authority to help me solve problems at work.” Cronbach’s α for leader member exchange scale was 0.896.
Ethical consideration
The Ethics Committee of Henan Provincial Key Laboratory of Psychology and Behavior approved this study (No. 20230402003). All participants were informed that participation in the study was voluntary and that they could withdraw at any time. Each participant gave their informed consent before they took part in this study. All data were confidential and anonymous and were only accessed by the researchers.
Anzeige
Statistical analysis
Using SPSS 21.0, we conducted descriptive statistics, correlation analysis, and reliability analysis for each variable before using Mplus 8.3 to test for structural validity, common method bias, and the hypotheses.
Analyses and results
Common method deviation test
We took measures during survey development and distribution to reduce single-source bias [49]. Harman’s single factor test was used to test four variables at the individual level, and the results showed that the first principal component explanation rate was 36.8%, which is lower than the 40% threshold. In addition, by controlling unmeasured potential method factors, a common method variance (CMV) was introduced into the confirmatory factor analysis, and the same loading was set for each item on the common method factor to further test for common method variance [50]. The test results are shown in Table 1. The fitting index of the four-factor + CMV model in the sample is χ2(47) = 132.84, RMSEA = 0.105, SRMR = 0.101, CFI = 0.906, TLI = 0.869. Compared with the fitting index of the four-factor model, the fitting indexes RMSEA, CFI, TFI and SRMR all deteriorated, indicating that common method bias was not a serious issue in this study.
Table 1
Model Fit results for confirmatory factor analyses
Model | χ2 | df | χ2/ df | RMSEA | CFL | TLI | SRMR |
|---|---|---|---|---|---|---|---|
four-factor + CMV model | 132.84 | 47 | 2.826 | 0.105 | 0.906 | 0.869 | 0.101 |
four-factor model | 105.78 | 48 | 2.204 | 0.085 | 0.937 | 0.913 | 0.071 |
three-factor model | 205.46 | 51 | 4.029 | 0.135 | 0.832 | 0.782 | 0.089 |
two-factor model | 351.97 | 53 | 6.641 | 0.184 | 0.674 | 0.594 | 0.119 |
one-factor model | 409.85 | 54 | 7.590 | 0.199 | 0.612 | 0.526 | 0.127 |
Aggregation test
Given that the moderating variable, group voice climate, is a team-level variable, it is necessary to make aggregation test. Climate ratings for each team were created by averaging intrateam ratings, aggregation indices (e.g., intraclass correlations [ICCs], within-group interrater agreement [Rwg]) provided support for this procedure [51, 52]. We conducted a one-way analysis of variance on group voice climate and found that the group voice climate of different teams had significant differences (F = 2.345, p < 0.01), ICC(1) = 0.161, ICC(2) = 0.674 (as shown in Table 2). The mean and median of the Rwg were 0.984 and 0.986 respectively. These results indicate that the team scores on group voice climate have high consistency.
Table 2
Polymerization test of Group Voice Climate
Variance | df | F | MSB | MSW | ICC(1) | ICC(2) | |
|---|---|---|---|---|---|---|---|
Between group | 2403.684 | 26 | 2.345** | 92.449 | 39.426 | 0.161 | 0.674 |
Within group | 5519.586 | 140 | |||||
Total | 7923.269 | 166 |
Confirmatory factor analyses
Before testing the research hypothesis, confirmatory factor analysis was carried out on four variables: voice behavior, affective leadership, supervisor-subordinate Guanxi, and leader member exchange. As shown in Table 1, each fitting index of the four-factor model meets the minimum standard and is superior to other models, indicating good discriminative validity of the four variables. In addition, the combined reliability of each variable is greater than the standard value of 0.6, and the square root of the average variance extraction value is greater than the correlation coefficient with other variables. Therefore, the aggregate validity and reliability of all variables in the scale are good.
Descriptive statistics
As shown in the Table 3, affective leadership at T1 is positively correlated with supervisor-subordinate Guanxi at T1 (r = 0.467, p < 0.01) and is positively correlated with T1 leader member exchange (r = 0.676, p < 0.01). There was no significant correlation between T1 affective leadership and T2 voice behavior (r = 0.039, p > 0.05). The supervisor-subordinate Guanxi at T1 was positively correlated with T1 leader member exchange (r = 0.560, p < 0.01) and significantly positively correlated with T2 voice behavior (r = 0.205, p < 0.01). There was no significant correlation between T1 leader member exchange and T2 voice behavior (r = 0.071, p > 0.05).
Table 3
Means, Standard Deviations and intercorrelations among variables (n = 167)
Variables | M | SD | CR | 1 | 2 | 3 | 4 |
|---|---|---|---|---|---|---|---|
1. AL T1 | 51.359 | 7.035 | 0.962 | (0.684) | |||
2. SSG T1 | 19.419 | 4.222 | 0.774 | 0.467** | (0.380) | ||
3.LMX T1 | 28.443 | 4.476 | 0.902 | 0.676** | 0.450** | (0.569) | |
4.VB T2 | 38.443 | 7.052 | 0.960 | 0.039 | 0.205** | 0.071 | (0.709) |
5.GVC T1 | 47.898 | 3.805 | - | - | - | - | - |
Hypotheses testing
Since all variables involved in the mediation effect test belong to the individual level, we conducted a path analysis to test the hypothesis, and followed mediation analysis method proposed by Hayes (2017) [53].
As shown in the Table 4, the results of Model 2 show that affective leadership has no significant predictive effect on voice behavior (b = -0.055, SE = 0.100, p > 0.05), meaning that hypothesis 1 is not supported. In addition, the results of Model 1 show that affective leadership can significantly positively predict supervisor-subordinate Guanxi (b = 0.272, SE = 0.039, p < 0.001). When affective leadership and supervisor-subordinate Guanxi are placed in the regression equation at the same time, the results of Model 3 show that affective leadership has no significant predictive effect on voice behavior (b = -0.136, SE = 0.115, p > 0.05) and that supervisor-subordinate Guanxi can significantly positively predict voice behavior (b = 0.422, SE = 0.135, p < 0.01). This indicates that supervisor-subordinate Guanxi plays a mediating role in the relationship between affective leadership and voice behavior. In this study, the bootstrap method was used to verify the above mediating effect again, and the results showed that the effect size of affective leadership’s influence on voice behavior through supervisor-subordinate Guanxi was 0.115(SE = 0.041). The 95% confidence interval was [0.043, 0.207], and the confidence interval did not contain 0, lending support to hypothesis 2.
Table 4
Unstandardized estimates of the mediation path model
Variables | SSG T1 | VB T2 | VB T2 |
|---|---|---|---|
M1 | M2 | M3 | |
AL T1 | 0.272*** | -0.055 | -0.136 |
SSG T1 | 0.422** | ||
LMX T1 | 0.132 | 0.055 |
All focal variables reside at Level 1, except the cross-level moderator (group voice climate), which resides at Level 2. It underscored the need test hypotheses in a multilevel framework. Our centering strategy aligned analyses with recommendations from multilevel methodologists [54, 55]. We group-mean centered supervisor-subordinate Guanxi and grand-mean centered Level 2 variables (group voice climate). We estimated a two-level model with subordinates (Level 1) nested in workgroups led by a supervisor (Level 2). We tested our model using multilevel path analyses in Mplus (as shown in Table 5) [56]. The results of Model 4 show that the random slope coefficients of supervisor-subordinate Guanxi and voice behavior have significant differences with different group voice climates (γ = 0.047, SE = 0.021, p < 0.05), establishing the moderating effect of group voice climate. Additionally, a simple slopes test reveals that the relationship between supervisor-subordinate Guanxi and voice behavior is positive when group voice climate is high (γ = 0.254, SE = 0.095, p < 0.01). Conversely, the relationship between supervisor-subordinate Guanxi and voice behavior is not significant when group voice climate is low (γ = -0.103, SE = 0.127, p > 0.05), and the difference between the high and low was significant (difference = 0.357, SE = 0.160, p < 0.05). To further ascertain the moderating effect of group voice climate, we followed Aiken et al. (1991) to demonstrate the relationship between supervisor-subordinate Guanxi and voice behavior [57]. We then conducted a regression analysis of supervisor-subordinate Guanxi on voice behavior respectively, for which the results are shown in Fig. 2.
Table 5
Unstandardized estimates of the Cross-level Moderation Model
Variables | VB T2 | |
|---|---|---|
M4 | M5 | |
AL T1 | 0.041 | |
SSG T1 | 0.076 | 0.074 |
LMX T1 | 0.079 | 0.038 |
GVC T1 | 0.110 | 0.091 |
SSG*GVC | 0.047* | 0.047* |
Fig. 2
Cross-level Moderation Model of SSG
×
![]()
To test for cross-level moderated mediation in this study, we respectively calculate a standard deviation above group voice climate, a standard deviation below group voice climate, and the difference of the mediation effect of supervisor-subordinate Guanxi [58]. The results of these calculations are shown in Table 6. Under the conditions of a high group voice climate, the mediating effect of supervisor-subordinate Guanxi is 0.027(p < 0.05) and under the conditions of a low group voice climate, the mediating effect of supervisor-subordinate Guanxi is -0.011(p > 0.05). The difference between high and low effect values is 0.038(p < 0.05), indicating that the mediating effect of affective leadership on voice behavior through supervisor-subordinate Guanxi is moderated by group voice climate. In addition, R software was used to repeat sampling (20,000 times) by the Monte Carlo method to obtain confidence intervals for the estimated indirect effects [59, 60]. When the moderating variable is at a high level and a low level, the 95% confidence interval is [0.006, 0.054] and [-0.043, 0.018] respectively (see Figs. 3 and 4), and the 95% confidence interval for the high-low difference value is [0.004, 0.081], excluding 0 (see Fig. 5). These results also verify the moderating effect of group voice climate.
Table 6
Test of the Cross-level Moderated Moderation Model
Moderation Variable | Group Voice Climate | Effect | 95% LLCI | 95% ULCI |
|---|---|---|---|---|
SSG | high: M + SD | 0.027 | 0.006 | 0.054 |
low: M-SD | -0.011 | -0.043 | 0.018 | |
high-low difference value | 0.038 | 0.004 | 0.081 |
Fig. 3
Mediating effect of the moderation variable at high level
×
![]()
Fig. 4
Mediating effect of the moderation variable at low level
×
![]()
Fig. 5
Mediating effect of high and low differences
×
![]()
Discussion
Our exploration of the mediating effect of supervisor-subordinate Guanxi between affective leadership and the voice behavior of nurses found that affective leadership promotes a sense of belonging and personal responsibility in nurses. Having enjoyed the benefits of a positive supervisor-subordinate Guanxi with the head nurse, subordinate nurses are more inclined to return the favor by seizing opportunities to promote the development of the organization. These opportunities may include forwarding their ideas and suggestions to improve processes and practices in the organization. Such nurses tend to have higher levels of job satisfaction [61], obtain more “relationship resources” [62], and devote more energy to work [63]. Simply put, they are well placed to identify existing problems and opportunities in the workplace for which they may contribute ideas based on their own knowledge and experience.
We have also examined the boundary conditions of the indirect effect of supervisor-subordinate Guanxi on the relationship between affective leadership and the voice behavior of nurses. We found that group voice climate can not only moderate the relationship between supervisor-subordinate Guanxi and nurses’ voice behavior, but also moderate the indirect effect of supervisor-subordinate Guanxi on the relationship between affective leadership and nurses’ voice behavior. Group voice climate, as a collective cognition of team members about voice behavior, will have a certain impact on the psychology and behavior of organization members. Previous studies have also found that group voice climate can give positive hints to team members, thus affecting their voice behavior [43, 46, 47, 64], but the group voice climate of the nursing community is understudied.
We found that when members of a nursing team perceive the team environment to be a safe space to speak out, that is, there is a high perception of group voice climate, these individuals are more inclined to voice themselves, particularly if they have enjoyed a positive supervisor-subordinate Guanxi with the head nurse. Generally, subordinates need to believe that there is a chance that their ideas and suggestions will be given due consideration by their supervisors if they are going to speak up. To foster an environment where subordinates feel encouraged to proactively present suggestions for addressing organizational challenges and identifying areas for improvement, it is imperative that supervisors not only demonstrate genuine care and concern for their subordinates but also remain attuned to their emotional states and address and alleviate negative emotion, by doing so, supervisors can cultivate a positive supervisor-subordinate Guanxi with subordinate. This relational dynamic is further reinforced by the reciprocity principle inherent in social exchange theory, whereby subordinates are more likely to engage in open communication and contribute constructive suggestions that benefit the organization. In addition, the leader should also build a good team atmosphere, so that the subordinates think that their suggestions can be adopted by the supervisors, their suggestions can bring benefits to the organization, at this time, the subordinates are more likely to voice under the influence of supervisor-subordinate Guanxi.
Limitations and future research
Our study adopted the method of convenient sampling, with participants hailing from the same region. To increase the universality of our results in future studies, we suggest that a probability sampling design could be used in different regions and cultural backgrounds. We also sought to reduce the issue of common method bias by undertaking data collection at two time points and through matched responses between supervisors and subordinates. However, our paper does not discuss the mutual influence between variables and the growth trend of individuals. Future studies may consider using tracking research, experience sampling, and other methods to explore this effect [65]. In addition, future study can also consider supplementing the analysis of qualitative data, combining quantitative research with qualitative research to explore whether the voice behavior fluctuates over time. While the method of a questionnaire survey served to address the research questions of our study, other types of studies, including situational experiments or even laboratory experiments, may also be helpful in providing further insights on the mechanism of voice behavior.
Acknowledgements
The authors acknowledge the participants for their acceptance to participate in the study.
Declarations
Ethics approval and consent to participate
The studies involving human participants were reviewed and approved by the Ethics Committee of Henan Provincial Key Laboratory of Psychology and Behavior (No. 20230402003). The participants provided their written informed consent to participate in this study. All methods in this study were carried out in accordance with relevant guidelines and regulations according to the Declaration of Helsinki.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
